A strategist who builds with teams, turning ambiguous AI and data problems into production systems that improve real decisions.
Selected Work
systems & research, 2023 → 2026
Built a cultural intelligence system that turns fragmented international signals into country rankings, corridor analysis, and decision-ready briefs.
I defined the strategic question and the measurement model, built the data pipeline and the product on top of it, then added a tool-calling analyst grounded in the live store.
The work continues in deployment: watching people use it to make decisions, finding where the system fails them, and shipping the fix.
stack python · fastapi · react · astro · cloudflarestatus in active development, ask for a walkthrough
problemwhich cultures are gaining ground, where, and how would you know. no measurement framework existed
usersstrategists and marketers inside a streaming and cultural analytics engagement
interventionmeasurement model, data pipeline, analytical product, tool-calling analyst
outcomequestions that needed a bespoke study now run against a live store, with the evidence attached
Built a conversational analyst that uses 14 tools against a live cultural data store. It streams answers, exposes the tools it used, and keeps claims traceable to the product corpus.
This gives analysts a natural-language interface to live evidence without relying on model memory. Tool calling over a structured store rather than embedding search, so every answer resolves to a result someone can go and check.
stack python · fastapi · sse streaming · tool-calling foundation modelstatus live inside the suite
problempeople could read the numbers but could not interrogate them without an analyst in the room
usersstrategists and analysts working inside the suite
interventionstreaming tool loop, 14 tools over the live store, tool disclosure in the answer
outcomelive evidence is queryable in plain language, and every claim traces back to a tool result
MSc research on intellectual monopoly theory and the institutional conditions that shape dominance in emerging technology markets.
An original framework tested with mediated and moderated regression, asking why some early technology markets consolidate into monopoly positions while others stay contested, and what that means for anyone trying to enter one.
method mediated & moderated regressionfield innovation economics, erasmus university
problemwhy some early technology markets consolidate into monopoly while others stay contested
interventionan original framework, tested on the data rather than argued
outcomegraded 9.0 for academic rigour
Independent research on Gen Z, social media, and climate-change content for a global communications organization.
End-to-end ownership: research design, trend analysis of digital advocacy, a 46-page report, and a final presentation of findings to the global leadership team.
output 46-page report + leadership briefing
problemwhy climate messaging was landing badly with an audience that already agreed with it
usersthe global leadership team of a communications organization
interventionresearch design, trend analysis of digital advocacy, a 46-page report
outcomefindings presented to global leadership as the basis for their approach
Culturemotif
live · culturemotif.com
Culture moves. We map it.
Culturemotif is the public product layer of my cultural intelligence work. It turns original analysis into clear stories about the cities and scenes gaining creative influence. Here is what I built:
the data
the research, the analysis, the claims underneath every story
the taste
the design system, the visual motifs, the assets
the structure
the front end, the infrastructure, the deploy
It shows how I work: one strategic question carried from evidence, to software, to an experience people use.
roleresearch & analysis · design system · front-end build · infrastructure · deploystackastro 5 · react islands · tailwind 4 · mdx · cloudflare pages & workerslive11 stories, a product page, a tenets page, generated og art and brand exports
About
the short version
I frame the problem, build the system, and ship the front end it arrives on.
I work where innovation strategy, data systems, and applied AI meet. Most of it starts with a question that has no measurement framework yet and ends with a working system that answers it: pipeline, analysis, output. Messy data is fine. So is a problem that turns up without a brief.
Forward deployed work sits between discovery, strategy, engineering, and adoption. I am strongest when the problem matters but the specification is still unclear. I frame the question, design the evidence, build the AI or data system, and then iterate with the people using it.
nowfreelance · building culturemotif and agentic systems · rotterdamfocusapplied ai × data systems × innovation strategytoolkitpython · fastapi · react · rag & retrieval · tool calling · aws bedrock · sagemaker · data pipelines · econometric modellingacrosssingapore · london · barcelona · new york · netherlandseducationmsc management of innovation, erasmus university rotterdam · bsc international business, university of twenteopen toforward deployed engagements · data strategy · agentic systems work
Contact
usually replies same day
Have an important problem without a clean spec?
I am open to forward deployed AI, data product, and strategy-to-build engagements.